how to use YAML for extra fields?

[mrswats]

In the readme you mention something about embedded yaml and extra fields, but I cannot see anything documented about it. How does it work? What do I need? I would love to have more data in my database instead of having to alter the generated database directly.

I love this tool, thank you!

[jgranduel]

Hi,
if I may, as I've just discovered this tool, do you mean something like described in pandoc documentation and reported in this short example?

file : a.md, containing embedded YAML for title, author, etc.

---
title: Test title
date: 2021-08-23
author:
    - Bob
    - Alice
    - And and some other authors
some_key: |
    and some very
    long multiline
    value.
tree_value:
    sub_tree : 
        - foo
        - bar
        - baz 
---

# title

## subtitle 1 

para 1.

## subtitle 2

para 2.

then

> markdown-to-sqlite.exe md.sqlite markdown_table a.md

.schema
CREATE TABLE [markdown_table] (
   [_id] TEXT PRIMARY KEY,
   [_path] TEXT,
   [text] TEXT,
   [html] TEXT,
   [title] TEXT,
   [date] TEXT,
   [author] TEXT,
   [some_key] TEXT,
   [tree_value] TEXT
);
.mode line
select * from markdown_table;
      _id = 5d48a79a3e66a0b67d03c582501493d256004f1f
     _path = a.md
      text = # title

## subtitle 1

para 1.

## subtitle 2

para 2.
      html = <h1>title</h1>
<h2>subtitle 1</h2>
<p>para 1.</p>
<h2>subtitle 2</h2>
<p>para 2.</p>
     title = Test title
      date = 2021-08-23
    author = ["Bob", "Alice", "And and some other authors"]
  some_key = and some very
long multiline
value.

tree_value = {"sub_tree": ["foo", "bar", "baz"]}

yaml metadata are serialized as JSON.

HTH

[mrswats]

Yeah, pretty much this is what I was not getting, many thanks!

It wasn't clear to me how I could: First, embed YAML into the markdown, and second, what would the result of this would be. It wouldn't have occurred to me to look at the Pandoc documentation!

This all makes a lot more sense now.

[mrswats]

Would it be worth it to open a PR to add this information in the README?

[jgranduel]

Happy I could help you.
AFAIK, metadata block should be terminated by --- but not ... as pandoc allows.

By the way, have you come across this post or SQLite documentation for automatically extracting data from json which is pretty relevant here. I changed the previous SQLite schema:

SQLite> alter table markdown_table add column extra_info generated always as (json_extract(tree_value, '$.extra_info')) virtual;

I created another markdown (b.md):

---
title: Test 2
date: 2021-08-23
author:
    - me
    - etc
some_key: |
    and some very
    long multiline
    value.
tree_value:
    extra_info: 
        "I am an extra info!"
---

# title b

## subtitle b  

para b.

> markdown-to-sqlite.exe md.sqlite markdown_table b.md

then sqlite3 md.sqlite (note the extra_info column):

sqlite>.mode line
sqlite> select * from markdown_table where title = 'Test 2';
       _id = bb801cb4e571ec0bc7a6ad6d316dacf27c7b107a
     _path = b.md
      text = # title b

## subtitle b

para b.
      html = <h1>title b</h1>
<h2>subtitle b</h2>
<p>para b.</p>
     title = Test 2
      date = 2021-08-23
    author = ["me", "etc"]
  some_key = and some very
long multiline
value.

tree_value = {"extra_info": "I am an extra info!"}
extra_info = I am an extra info!

And all the thanks to Simon Willison!

[mrswats]

Opened #6